Final report for GW17-059
Project Information
Livestock production is important to the global economic and food security (Thornton et al., 2009; Thornton, 2010). The United States beef industry alone had a retail equivalent value of approximately $105 billion in 2015 according to the United States Department of Agriculture Economic Research Service Cattle and Beef Statistics and Information (USDA 2017). The provision of range and pasture forage is one of the most practical ways to meet the needs of cattle, and ranchers use many different grazing management systems to avoid risk and meet their individual operation objectives (Quaas et al., 2007; Roche et. al, 2015). The animals utilizing these rangelands are subject to seasonal shifts in forage quality (Kunkle et al., 2000; Cline et al., 2009; Jung et al., 1985). Coupled with seasonal declines in forage quality is the impact climate change has on forage quality and quantity. Augustine et al. (2018) suggests that climate change will result in a decline in overall forage quality, specifically nitrogen content of palatable species, and a 38% increase in forage quantity. However even with a quantity increase in forage, climate change is projected to have a negative effect on livestock production by requiring supplementation earlier in the grazing season to counteract the declines in forage quality (Augustine et al., 2018). Understanding what forage species are being utilized by cattle during certain parts of the year may help ranchers develop better grazing strategies and aid them in creating supplementation plans during critical time periods.
The overall goals of this study were to 1) validate DNA metabarcoding with a feeding trial, and 2) create a snapshot of cattle diets across the state of Wyoming while introducing a new technology to the general public.
Cooperators
Research
Rancher Outreach
For the diet survey, rancher involvement was gained mainly through the University of Wyoming Extension and word of mouth of participating ranchers. Overall, we had 31 ranches from 20 counties in Wyoming and 2 ranches in Nebraska participate in the 2017 ranch diet survey. Once signed up, the ranchers (or their respective Extension agents) were mailed kits which included three sampling bags for June, July, and August, a cooler, an ice pack, and prepaid shipping back to the ARS location. For each sample, ranchers were instructed to subsample ten individual, fresh fecal pats (approximately 1 tablespoon of material each) and combine in the Ziploc bag for the corresponding month. They were then instructed to freeze the sample until shipping it to the ARS location. Once received, all samples were thawed, processed, and shipped to the Jonah Ventures Laboratory in Boulder, Colorado for the DNA metabarcoding analysis. At the lab, samples were sequenced and diet composition results were obtained for each operation. A technician combed through the dataset using the USDA Plants Database to identify species located in Wyoming and discard non-sensical OTUs. Individual reports were created using Piktochart to send to each ranch with the top 7 species contributing to the dietary protein for each individual herd and each month of sampling. Raw Excel files were also sent to individual ranches.
Upon receiving results, ranchers or extension agents reached out to help understand the results, the process undergone to clean-up the data prior to sending results out to them, and to learn the limitations of the technology. Because DNA metabarcoding uses the UAA intron of the plant chloroplast DNA to obtain species composition, the technology is weighted by the protein content of the plants. Thus plants with higher protein contents, such as legumes, often have higher percentages than plants with lower protein contents. Rather than detecting actual percentage of mass consumed, this technology detects the contribution of specific species to the overall dietary protein content of the diet (Craine et al. 2016). It is possible to reverse engineer the mass consumed if you know the protein content of the individual plant species at the specific point in time samples were taken. However, since protein is often limiting for ruminant animals, it can actually be helpful to understand the dietary protein contribution of individual plant species. To better model influences on animal diet, we obtained metadata from the Oregon State PRISM Data Explorer using the latitude and longitude of towns nearest the individual ranches. We conducted some preliminary analyses to understand what cattle diets look like across the state of Wyoming during the summer grazing season and the main drivers in cattle diet, but results are inconclusive at this point in time.
Feed Trial
To test the validity of DNA metabarcoding analysis, we conducted a feeding trial in February-March of 2017. Five non-gestating and non-lactating 2 year old heifers were used in a six week feed study. Animal care and use was approved under Institutional Animal Care and Use Committee (IACUC) protocol # 20170208DS00258-01. Heifers were individually penned in adjoining 160 m2 pens that included shade, outdoor access, automatic water, and were sloped to promote drainage away from feed bunks. Heifers were weighed prior to the feeding trials and ranged from 416 to 527 kg. For homogeneous diets (1, 2, and 6) animals were fed ad libitum. Heterogeneous diets were fed at metabolic weight (MW) based rations using MW0.75 based on heifer weights. Each diet was fed for one week and fecal samples were collected from each heifer on day 7 prior to switching to the next diet. All samples were immediately frozen after sampling and sent to the Jonah Ventures lab for analysis. The six diets were: (1) C3 grass – creeping meadow foxtail (Alopecurus arundinaceus Poir.), (2) C4 grass – foxtail millet (Setaria italica (L.) P. Beauv.), (3) C3 grass /C4 grass /alfalfa (Medicago sativa (L.)) equal proportion, (4) C3 grass/C4 grass/alfalfa equal proportion + minor component of Wyoming big sagebrush (Artemisia tridentata Nutt. ssp. wyomingensis Beetle & Young) leaves, (5) Alfalfa + minor components of crested wheatgrass (Agropyron cristatum (L.) Gaertn.), western wheatgrass (Pascopyrum smithii (Rydb.) Á. Löve), and blue grama (Bouteloua gracilis (Willd. ex Kunth) Lag. ex Griffiths), and (6) alfalfa. Minor components were fed at 2.0 grams per day for blue grama and 12.6 grams per day for crested wheatgrass. Due to technician error, western wheatgrass was fed at 12.6 grams on day one of the minor component addition, fed at 3.0 grams day two and three, and 4.6 – 6.0 grams day four. Across all four days, the animals consumed an average of 5.8 – 6.15 grams daily of western wheatgrass.
DNA metabarcoding results were organized into three different groups: 1) ‘as-fed’, 2) ‘laboratory non-discretionary’, 3) ‘laboratory rectified’. The ‘as-fed’ diet represents the expected results using the crude protein contents of the individual components fed at that point in time and the expected proportions we should have seen of each species in the results. The ‘laboratory non-discretionary’ represents the results from the lab without technician discretion in grouping highly related species or using botanical knowledge to remove extraneous Operational Taxonomic Units (OTUs). The ‘laboratory rectified’ diet results use the technician’s knowledge of the plant community and the technology to identify species that are highly related and to discard plant species not found in the U.S. using the United States Department of Agriculture (USDA) Plants Database. Though the technology was able to detect major diet components, it did not detect most of the minor components in the correct proportions. We conclude that, it is necessary for individuals to know the plant community they are working in to obtain accurate results when using this technology on free-ranging animals.
Table 1. Date of feeding trial initiation and diet composition to validation DNA metabarcoding of fecal samples for predicting plant species composition of cattle diets.
|
Start Date |
Fed Diet |
|
2/7/2017 |
1st – C3 Grass Uniform (ad libitum) |
|
2/14/2017 |
2nd – C4 Grass Uniform (ad libitum) |
|
2/21/2017 |
3rd – C3 + C4 + Legume (metabolic weight based ration) |
|
2/28/2017 |
4th – C3 + C4 + Legume + Minor Component of Sagebrush leaves (metabolic weight based ration) |
|
3/7/2017 |
5th – Legume + Minor components of Agropyron cristatum + Pascopyrum smithii + Bouteloua gracilis (metabolic weight based ration) |
|
3/14/2017 |
6th – Legume Uniform (ab libitum) |
Ranchers Outreach
We conducted a PCA and CCA on the dataset obtained for all ranches across the state of Wyoming and portions of Nebraska; however, results were unclear and did not explain the variation in the dataset at a satisfactory level. Thus our results are inconclusive at this point in time.
Feed Trial
Our results indicate that this technology does a decent job of detecting major components in the diet; however, minor components are not as easily or accurately detected. Within every dataset over 2000 individual OTUs (representing a suite of species with a 97% base pair matching of the targeted gene sequence) were found. However, in some diets only a few individual plant species were fed. Thus it is important to discard extraneous OTUs at the tail end of the dataset to avoid identifying minor components that are not actually part of the diet. For our purposes, the top 100 OTUs were identified and used for analysis.
Research Outcomes
Education and Outreach
Some ranchers have expressed interest in a webinar or workshop discussing how the methodology works so we are hoping to do this in the future.
Participation Summary:
Thus far, most of our outreach has been via email or during small workshops put on by other organizations. We are hoping to conduct a workshop/presentation of how this technology works in the near future. I have discussed the methodology at 2 range workshops. We have sent most individual operations their results. Interestingly, some areas are detecting invasive species in cattle diet that were not thought to be widespread in the areas sampled (although the species was known to be in the county). There is potential that this technology may be used to help detect the presence of invasive species and allow for early containment of said species.